TWI394197B - A process for controlling roximity effect correction in an electron beam lithography system - Google Patents

Description

Method for controlling proximity effect correction in an electron beam lithography system

The present invention relates to a method for controlling proximity effect correction in an electron beam lithography system. The method is suitable for accurate numerical determination of neighboring parameters of a point spread function (PSF) for optimally controlling proximity correction in high resolution electron beam lithography (EBL).

The proximity effect parameter is a dedicated numerical input that controls any proximity effect correction software. Etching the shaped beam to meet the high critical dimension control "CD control" requirements (depending on the actual international semiconductor technology route ITRS of International SEMATECH) and compensating for Gaussian functions associated with subsequent process steps (development, etching, etc.) The reticle of the shaped beam and/or the pattern deviation in direct writing processing.

A number of methods have been proposed to determine proximity parameters that reflect various effects. In addition to the proximity effect, an atomization effect occurs simultaneously in the electron beam lithography system. Below are a few published literature on the application of proximity effect correction.

In the article entitled "Optimum PEC Conditions Under Resist Heating Effect Reduction for 90 nm Node Mask Writing", Proc. SPIE, Vol. 4889, Part 2, pp. 792-799 (paper No. 86), The critical dimension (CD) change caused by 50KV electron beam writing, the problem of resist heating and proximity effect, the experimental method is determined by the adjacent input parameters in the reticle process, and the experimental method adopts a stepwise change separately. A large area matrix of adjacent corrected test patterns written under various conditions of adjacent parameters. In this way, the optimal parameters are determined from the direct measurements on the test patterns with the least graphical distortion effects. This experiment and the evaluation of this image are very time consuming. Since the number of possible combinations of input parameters is large, the method is limited to the Gaussian function approximation of the last 2 PSFs obtained. This method is widely used in the production of photomasks.

See Microelectronic Engineering 5 (1986) 141-159, article entitled "Determination of Proximity Parameters in Electron Beam Lithography Using Doughnut-Structures" in North Holland. The test structure is a ring that is used to determine the parameters in the correction function. This method provides an easy way to determine proximity parameters from an exposed array of annular bodies by means of an optical microscope. This method is not sensitive enough to obtain CD control using an electron beam, and is not suitable for the EBL method of high resolution patterns.

In the article "Point Exposure Distribution Measurements for Proximity Correction Electron Beam Lithography on a sub-100 nm Scale" by J. Vac. Sci. Technol. B5 (1), Jan/Feb 1987, a single dot/pixel is exposed. In a wide range of doses, the measured pattern diameter and results directly approximate the Gaussian function. This method can only be applied to special high contrast resists (ie, insensitive to changes in development effects), requires high resolution measurement techniques (SEM), and additional methods (graphic stripping or deposition) Coating). This method may not be applicable to a large number of chemically amplified resists (CARs) used. With very high dose point exposure methods, the acid diffusion effect may exceed the true nature of the proximity effect [Z. Cui, PhD, Prewett "Proximity Correction of Chemically Amplified Resist for Electron Beam Lithography," Microelectronic Engineering 41/42 (1998) ) 183-186].

The article "Determination of Proximity Effect Parameters in Electron-Beam Lithography" in J. Appl. Phys. 68 (12), 15 December 1990 discloses a method for determining from a rectangular array of mesh patterns in electron beam production. The experiment is based on a method according to which the data will be corrected by means of illumination optical inspection after the data has been processed adjacent parameters. The test pattern to be measured is used to determine the proximity effect. This method is not suitable for modern conventional high resolution product electron beam lithography systems.

The atomization effect is also referred to in some publications. The article "Fogging Effect Considerationin Mask Process at 50KeV E-Beam Systems" proposes a proposal to reduce the atomization effect in high voltage electron beam systems.

An article in Microelectronic Engineering 5 (1986) 141-159, North Holland, entitled "Determination of the Proximity Parameters in Electron Beam Lithography Using Doughnut-Structures", and at J. Appl. Phys. 68(12), 15 December The article "Determination of Proximity Effect Parameters in Electron-Beam Lithography" in 1990 reveals the atomization effect.

It is an object of the present invention to achieve a method by which the brightness parameters of an electron beam lithography system can be reliably corrected by studying the effects of the atomization effect.

This object is achieved by a method according to claim 1 of the scope of the patent application.

A method for controlling the proximity effect correction in an electron beam lithography system achieved the above object, wherein the exposure is controlled in order to obtain a consistent pattern with the design information obtained after the final processing, the method comprising the steps of: ^* Exposing an arbitrary set of graphics without applying the method for controlling proximity correction; ^* measuring the geometry of the resulting test structure and thereby obtaining data of a set of measurements; ^{* determining} from the data of the set of measurements The adjacent input ranges of the basic input parameters α, β and η are numerically represented; ^* by changing at least the basic input parameters α, β and η of the control function to the data set of the measurement results, and thereby obtaining the optimized parameter set, Adapt a model, ^{* Use} the correction function for exposure control of the electron beam lithography system during pattern exposure according to the design data.

In addition, it is useful to use the determined set of neighboring parameters for a calculation and compare the result to a set of measurement data for transparent and opaque lines that are "on target" with a slight dose exposure. Another possibility is to apply the adapted set of neighboring parameters to a calculation, and compare the result with a set of measurement data from other graphs of other cone-like graphs, and to compare the result with the representative points from the test graph. The measured data of the measurement results are compared. Yet another possibility is that the adapted set of neighboring parameters is applied to a calculation, and the result is compared with a set of measurement data from other arbitrary patterns of duty lines (duty-ratio), and the result is The measurement data of the measurement results in the test pattern representative points are compared.

The method is based on the analysis of graphical geometric variations as a direct response to non-corrected and/or interacting non-corrected graphics in EBL (electron energy, resist material, substrate material, front and back exposure processes) Process, graphics transfer, etc.). Back simulation is used to reconstruct the measured pattern change characteristics by introducing dedicated proximity parameters into the model. The data calculated from the model represents the lateral shape position of the analog image at the same point measured on the real graph. The measured data is compared with the calculated results at the same point on the representative test pattern (individual transparent/opaque line segments, pyramid-like patterns, linear arrays in the duty factor, etc.), and the determined adjacent parameter sets are visually observed. quality.

In this case, the required correction algorithm to work under the same model concept as used for the model will be satisfied, and the method also predicts possible graphical uniformity errors (graphic consistency) and in proximity correction, The resolution range after the actually determined neighboring parameters is used.

An advantage of the present invention is that it has an interpretation of the original geometric deformation using an analysis based on the model and an uncorrected representative image that is exposed (analysis of the direct process response as a typical graphical geometric change), the graphic being specific It was measured at the point (using a commercial measurement industry such as CD-SEM) and the data of the subsequent processing was recorded. Successive "back simulation" processes are the best possible reconstructions to be used for these effects. “Back Simulation” refers to the best approximation of the geometrical change measurements for specific graphic details based on pre- and post-exposure conditions and/or proximity (around graphic size) effects (=graphics and process reconstruction). A good calculation method for the number of input parameter sets. Once such graphical detail can be a change in the size of the pattern at a particular point as a function of exposure intensity (eg, in the simplest example, the change in line width and/or contact size versus exposure dose in two tones) ). Another variation may be, for example, the position of the adjacent pattern (eg, the relationship of the line width measurement to the change in the gap width of the large flange, a cone-like pattern, and/or the line one in the grid in the duty factor) Line). As a result, after introducing the obtained parameters into the model, a suitable simulation shows the same tendency to change the correlation of the geometric images obtained from the measurements. Thus, if a high speed algorithm works as the same model concept used in the model, then it results in a good recovery of the additional deformation effects of these input parameter sets in the proximity correction.

Measurement and simulation up to the minimum resolvable pattern size is achieved, which also allows forward scattering by electrons, secondary electron distribution, beam blur, resist effects (development, acid diffusion, quenching) and pattern transfer (microloading) The precise determination of the parameters described by the known "short-range" effect. As a result, proximity corrections that work with this parameter set will also work correctly in deeper sub-100 nm lithography intersections.

Experimental measurements on a pair of exposed patterns (described in the appendix "Test Patterns") are provided to provide all the necessary values to PROX-In (PROX-In is built to help determine the proximity effect parameters for the etcher) The function of the software tool above the user's smooth Window ^{T M} ) has no active editing dialog box and creates a premise of a simple ASC11-file containing the test data. These data then serve to determine the basis for selecting a particular built-in algorithm for the proximity parameters in the program. In order to minimize the degradation or distortion of the pattern due to submicron characteristics, it is inevitable to use a correction method to handle this effect. The prior art relies on: a) successive adjustments of the exposure dose, b) correction of the geometric image of the graphic, or c) a combination of the two methods mentioned above.

The main advantage of this method is that it does not require the use of large matrices of exposed adjacent correction patterns with various input parameters. The parameters will be determined from the measurements on a non-corrected simple test pattern. The amount of data and/or parameters to be analyzed is greatly reduced. The advantages of the present invention are as follows. The present invention uses only a relatively simple combination of a small number of exposed test patterns. The area of the substrate (5 inches and larger) covered by the test pattern was limited to 1% below. Moreover, the test pattern is exposed without any proximity correction. In addition, there is a possibility to change the overall graphics loading by the base "model" exposure of the additional auxiliary graphics around the test pattern. This can determine the change in pattern loading according to deviations during development and/or etching. There is a tendency to directly observe the deterioration of the pattern by individually changing one of the input parameters. There is then an interaction fine tuning of the input parameters to obtain the best possible CD-requirement (CD linearity). Using two or more Gaussian input parameter sets (Gaussian functions) with direct test possibilities, where and why additional Gaussian functions with various parameters are needed can achieve better results. The back-simulation and re-construction of the specific graphical details of any adjacent parameter sets allows for predictable changes in the parameters of the set of parameters for the given geometric combinations in the CD.

Under the real conditions of the product, the computer program "PROX-In" for the optimization and testing purposes of the method described in this application was developed and implemented.

Fig. 1 is a block diagram showing an electron beam lithography system 1. The electron beam lithography system 1 has an electron beam source 2 that emits an electron beam 3. This specification only mentions the electron beam 3. However, it must be understood that the invention is not limited to electron beams. In general, the invention can be used with particle beams that can be applied to the writing of pattern 5 on substrate 4. The substrate 4 itself is placed on a stage 6, which can be moved by the motors 7 and 8 in a plane spanning the X-coordinate X and the Y-coordinate Y. After exiting the electron beam source 2, the electron beam 3 is directed through the beam to the coil 9. After the beam is aligned with the coil 9, in the direction of propagation of the electron beam 3, the beam blanking unit 10 is provided. After that, the electron beam 3 reaches the magnetic deflection unit 11, which generally comprises four magnetic coils 12. After that, the electron beam 3 is directed to the substrate 4. As already mentioned, the substrate 4 is placed on the stage 6. The actual position of this stage is controlled by position feedback means 13. In addition, the electron detector 14 is placed in the vicinity of the stage 6. A computer 15 is provided for controlling the entire electron beam lithography system 1. In particular, the beam parameters are controlled, measured and adjusted to produce a pattern of constant size. The computer 15 is coupled via an interface 16 to an electron beam lithography system 1, which implements analog to numerical and/or numerical to analog conversion. The interface 16 is connected to the beam obscuring unit 10, the magnetic deflection unit 11, the position feedback device 13, the electron detector 14, and the motors 7 and 8 of the moving stage 6. Through the display screen 17, the user can obtain information about the setting and/or adjustment parameters of the electron beam lithography system 1.

Figure 2a is an example of a pattern 20 for covering a certain area 21, and the area 21 is filled with a plurality of Gaussian beams 22. Each Gaussian beam has the same diameter. In Fig. 2b, the shape of the cross section 23 of the Gaussian beam 22 is shown. A plurality of beams cover the area 21 required for the pattern 20.

Figure 3a shows an example for a graphic 30 that is written with a shaped beam 32. The total area 31 of the graphic 30 is covered by a plurality of shapes of the variable shape. The variable shape graphic fills the area of the graphic 31 to be written. In this case, the region 31 is covered by three different shapes 32 ₁ , 32 ₂ and 32 _{3 of} the electron beam. Figure 3b shows the shape of the cross-section 33 of the shaped beam 32, wherein the shape of the individual beams can be adjusted according to the pattern to be written. As shown in Figure 3b, the shape of the beam can be changed. This is indicated by arrow 34.

In both cases (Gaussian beam or shaped beam), submicron properties or graphics become a critical issue in reticle writing. With this pattern size, the electron beam lithography system faces a common parasitic electron scattering effect that causes unwanted exposure deposition in the surrounding area of the pattern being written. This parasitic electron scattering effect is called proximity efficiency [see, for example, T.H. Chang. "Proximity effect in electron beam lithography," J. Vac. Sci. Technol 12 (1975) p. 1271]. In the case where the smallest feature size becomes smaller than the backscattering range of the electron, the graphic overlay affects the size control of the graphic to be written. On the other hand, front scatter limits the maximum resolution. As the electron energy increases, the difference between the back and forward scatter increases. Any graphic detail belonging to a dedicated area is significantly distorted in the image of the resulting lithographic pattern compared to its original design size and shape. To minimize the degradation/deformation of a submicron-like pattern, it is inevitable to apply a correction method to deal with this effect.

Figure 4a shows a simulated orbit 42 for 100 electron scattering in a polymethyl methacrylate layer 40 (PMMA) on a GaAs substrate 41 defining a resist. The initial energy of the electron is set to 15KeV. When the electron beam 43 exits in the thin layer of PMMA, the electrons scatter and move according to the calculated orbit. Figure 4b shows a simulated orbital scattering of 100 electrons in a thin layer of PMMA plated on a GaAs substrate 41, wherein the initial energy of the electrons is higher as calculated in Figure 4a. In electron beam lithography, the main deformation is due to the electronic interaction of the resist/substrate system with additional effects that cannot be exactly separated and processed separately. Here, the main task is the absorbed energy density distribution (AEDD) dispersed in the resist, with a corresponding irradiation-chemical event distribution in the resist volume, creating a latent image in the resist (resistance Variation of the agent). The AEDD model in the resist layer is possible by using statistical (Monte Carlo) or analytical (transport equation) calculations of the electron scattering process. After absorbing the necessary radiation quantum from the exposure, a true potential image is formed by the local chemical change of the irradiated resist volume.

Figure 5a shows a schematic diagram of the form of the pattern 50 that needs to be written in the resist on the substrate. The graphic 50 has four different components that are clearly described as their final dimensions from the graphic design data. The first member 51 is a straight line having a defined width. The second member 52 is a rectangular shaped region. A straight line extends from the upper corner and forms the lower corner of the joint. The third member 53 is a rectangular shaped region. A straight line extends to the left to form the lower corner of the area. The fourth component 54 includes two regions that are connected by a straight line at their lower corners. An additional straight line extends from the upper corner of the area to the left.

Figure 5b shows the resulting pattern 55 being written in the resist without the correction according to the invention applied. The proximity effect in the resist layer becomes clearly visible. In the first position 56, the two regions are separated by a line where significant line width widening and bending of the two region shapes occur. There is no longer a gap between the area and the line. At the second position 57, where the two joint regions face each other, the deformation results in an interconnection between the two joint regions.

Figure 6 shows the first possible test pattern 60 (which has been implemented in PROX-In) being written into the resist. The process of simulating this first test pattern 60 is referred to as a tapered "PYR" (using a cone-like test pattern 60). This special method can be initiated by the user via the PROX-In user interface (see Figure 8). This process can determine the input parameters after analyzing the experimentally measured data from the line width variation of the exposed symmetric cone test pattern 60. The first test pattern includes lines 61 having a predefined line width 63. On both sides of the predefined line 61, along the line 61 being measured, a large flange 62 having a variable gap width 64 is exposed. In the case of non-correction, a predefined single clear line width 63 increases as the gap width 64 decreases between the measured line 61 and the plurality of pads 62. The measurement is made at the position marked with a dot 65 in Fig. 6. The dependence of the designed gap width on the measured line width 63 is the basis for the inputs obtained for the calculation and back simulation of the method (see Figure 9). The first line 91 of the obtained data includes the width of the gap that has been designed and exposed, indicated by [μm), while the second line 92 includes a suitable line width 63 of the exposed and tested line width. For this process it is necessary to find the optimum exposure dose for the non-corrected cone test pattern 60, here a single pre-defined exposed line (for example, in Figure 6, on the right side of the first test pattern 60) The lines above are not affected by large areas) and meet the target. The only first highest value shown in Fig. 9 is the start value (here, for example, 2 μm) of the line width 61 of the only large gap width 64.

Figure 7 shows a second possible test pattern 70 that has also been implemented in PROX-In, the PROX-In of which can be used to directly determine the proximity parameters. Similar to the other methods described above, the process is based on the measured value of the line width 71 of the exposed non-corrected duty factor test pattern "DRT". The plurality of lines 72 are exposed and/or otherwise processed in the resist. Lines 72 are formed in an array 74 of various spacings 73 between the lines. This method can be initiated by the user via a dedicated PROX-In user interface (see Figure 10). This process allows the proximity input parameters to be determined after analyzing the experimentally measured data representing the width variation of the line 75 from the exposed symmetric second test pattern 70. To receive the stored information (see Figure 11), the data in the two rows of the form is provided. The first line 111 is the duty factor of the values 1, 2, 3, ... 20 as ratios (1:1, 1:2, 1:3...1:20). The second row 112 is the measured line width in μm for a suitable ratio. The change in line width must be measured from somewhere in the center of each array 74 for various line/pitch ratios. The dot 75 in Fig. 7 indicates the position of the measurement. This is important prior to initiating the "DRT" measurement process, which is necessary to determine the optimal exposure dose for the single transparent line 76 on the right side of the second test pattern 70. In other words, the single transparent line that has been measured has the line width as required for the CAD-data, and the lines that have formed the pattern meet the target as best as possible.

Figure 8 shows an input window 80 for the user to initiate exposure of the first test pattern 60 as shown in Figure 6. The user selects the cone process by setting a mark (check "PYR" button) 81 above the "PYR" name, and then the back simulation process begins. The results are shown in Table 90 (see Figure 9). The measurement results obtained from the exposed first tapered test pattern 60 are set in the second (central) row 92. The first row 91 shows the gap width, while the third row 93 shows the line width 63 from the simulated back simulation/rebuild/calculation for a given set of input parameters. Figure 9a shows the results from PROX-In in curve form 94, the purpose here being to find such a parameter set 95 that provides the most suitable measurement data (black) 96 with calculated data (red) 97.

Figure 10 shows the PROX-In input window 100 for the user to initiate the process of the second "DRT" test pattern 70 as shown in Figure 7. The user selects the duty cycle test process by setting the mark 101 (checking the "DRT" button) above the "DRT" name, and then the calculation process begins. The results are shown in Table 110 (see Figure 11). The measurement results obtained from the exposed second test pattern 70 are set in three lines. The first row 111 contains data with a duty cycle, the second row 112 contains data with measured line widths, and the third row 113 contains line widths from back simulation calculations.

Figure 11a shows the results from PROX-In in the form of a graph 114, the purpose here being (also the same as in the case of a tapered pattern in the front) is to find such a provision of the measurement data 116 that best fits the calculation data 117. Parameter group 115.

Obviously, in order to obtain additional experimental and simulated data, other test patterns can be designed and matched, and for these data, the determined adjacent parameters must be cross-checked and made to match. This is consistent with providing a set of parameters that enable the micrograph to be exposed to the exposure, and the results obtained are highly consistent with the design data provided for the desired pattern, in other words: any pattern exposed by the method according to the invention results in a basis A graphic of the size required to design the material.

PROX-In operates on a standard computer 15. The computer 15 operates under Windows and does not require any dedicated hardware/software components. The installation of PROX-In is very easy. Create separate indexes and copy the supplied/provided files here. The general structure of the PROX-In is clear from the display main window 120 associated with the computer 15. After starting the program PROX-In, the main window appears immediately (see Figure 12). The main window 120 is divided into three main sections. The first portion 121 is in the entire upper half of the main window 120. The first part has the title of "calculate α" and "calculate β and η". The first portion 121 is composed of first, second, third and fourth separate sub-frames 121 ₁ , 121 ₂ , 121 ₃ and 121 ₄ . The first sub-frame 121 _{1 is} added with the title "α". The second sub-frame 121 _{2 is} added with the title "β-manual". The third sub-frame is titled "β-Auto". The fourth sub-frame is added with the title "η". Each of the sub-frames acts as the only first evaluation of the fast and lithographic processes from the relatively large graphics/wide lines of measurement responses and conveys the role of the method for approaching the first (general) value of the parameters.

The second portion 122 is placed at the bottom of the main window 120. The second portion 122 has a "simulated" heading thereon and serves as the final "fine tuning" of the parameters reconstructed from the best graphics using back simulation.

The third portion 123 is located on the right bottom side of the main window 120. The third portion 123 is a scroll-shaped "input/output" MEMO-box with a content window 124, where some necessary information resulting from the selection operation/calculation occurs.

The new electron beam lithography system is designed to meet the CD needs of the 100nm device processing level and below. In order to comply with these specifications, there must be a suitable knowledge base covering all graphical degradation/deformation effects through the entire process, as well as a coherent application of accurate calibration methods.

In the electron beam lithography system 1, the main variant originates from the electronic interaction of the resist/substrate system surrounding the additional effects, which are not exactly separate and separately processed. Here, the main task is the absorbed energy density distribution (AEDD) dispersed in the resist, the resist having a corresponding irradiation-chemical event distribution in the resist volume, and the resist volume is in the resist Potential image created in the etchant (resist variation). AEDD model simulation in the resist layer is possible by statistical (Monte Carlo) or analytical (transport equation) calculations using electron scattering processes. After absorbing the necessary amount of exposure from the exposure, a true potential image is formed by local chemical changes in the irradiated resist volume.

In general, the proximity correction control function f(r) is described as the sum of two or more Gaussian functions (see Equation 1). In the case of a standardized 2G-function, it is expressed as follows:

The first item α- represents the short-range characteristic of forward scatter, the second item β-backscatter, the parameter η- is the ratio of the backscattered component to the deposition energy of the forward scatter component, and γ- is the point of departure from the electron. Distance (see Figure 4a).

In a suitable developer, after applying a post-exposure process (mostly a wet process), a final de-resist mask can be obtained. For models and predictions of true resist pattern geometry, it is necessary to know exactly the solubility characteristics of the polymer that is altered by irradiation. The development process causes a large amount of uncertainty in the overall simulation due to the highly nonlinear nature of this complex thermo-hydrodynamic-dynamic process. Due to the existence of a wide variety of systems (semiconductor substrates, resists and post-exposure processes), the parameters in Equation 1 need to be determined for all of these different systems.

In the fabrication of the reticle, a similar complexity clearly indicates that the second step is also transferred to the imaging layer and/or substrate by a resist in both wet and/or dry etches.

Correction of proximity effects in the field of electron beam lithography systems is available through certain commercial package software, which are based on the principle of two or more Gaussian approximations using electron scattering images as described above. Handle the problem of optimizing the exposure dose. If the input parameters are only determined from Monte Carlo simulations, the calculations only involve pure AEDD. This result does not contain any information about additional nonlinear effects from other influencing factors. One influencing factor is this process, for example, irradiation-chemical events in the resist, thermal effects, and dissolution characteristics in development, such as etching in reticle fabrication. Another influencing factor is the tool, (for example, electro-optical aberrations and space charge effects that affect gas image slope and/or edge resolving power), relying on slamming to affect the resulting graphical distortion. With this in mind, the inputs used for the correction scheme should be estimated using an object property model that accurately describes these effects. The values of these parameters from the experimental measurements are also finely tuned.

For the calibration process, this system can only be operated if the value input is properly selected. Therefore, great efforts have been made to develop a fast and easy method for determining the numerical value of the process-related input parameter set required to determine the exposure correction algorithm. The flexible package PROX-In will help lithographers find/determine these optimized values.

Special care was taken to synchronize both the corrector's algorithm and the PROX-In software, and the same results were obtained for the same input parameters in the simulation mode. The present invention employs the concept of semi-portraits.

With the proximity correction of the electron beam lithography system 1, the size error on the reticle is reduced to <10 nm for devices producing 100 nm or less.

Before starting PROX-In, it is inevitable to extract the following main numerical values of lithography parameters directly from the combination of specially designed and exposed test patterns (for numerical calculations, import/set parameters to PROX-In). Required in the middle).

FIG. 13 shows an attached window 130 of the first portion 121 of the main window 120. In the first method, the alpha-parameter is calculated from the Monte Carlo calculation as a function of the actual electron energy and the resist material and thickness used. In the satellite window 130 labeled "Calculation a", the selection button "ALPHA(α)" 131 makes it possible to obtain a simple polymer material consisting of C, H, and O atoms (optional average density at The orientation alpha value of PMMA-(C ₅ H ₈ O ₂ ) between 1.1-1.3 [g/cm ³ ] (only related to pure electron elastic forward scattering, independent of the substrate material). The user has an input section 132 for the resist material, an input section 133 for electron energy and an input section 134 for the resist depth. The input of the required electron energy [KeV] and the depth [μm] in the resist at α should be calculated.

In general, it is not recommended to directly derive the alpha value as an input into the calibration process. This calculation is only close to the correlation between the exposure energy used and the thickness of the resist versus the change in alpha value. The true alpha final parameter will normally have a slightly larger value because of the additional process (development, etching) and/or tool-related errors (optical aberrations, current - in addition to forward scattering of electrons) , pre-, and post-exposure process stability and replication problems) will also affect this parameter.

For the actual process, the "back simulation" is used, and after a rough estimation of other parameters (β, η, ...), the final α-parameter determination can be completed.

Figure 14 shows an ancillary window 140 of the first portion of the main window 121 on the display screen or user interface for calculating β and η. The β-parameter represents a "long distance" collision with respect to the proximity effect of image distortion in an electron beam lithography system. This means that the β-value affects and finely determines the interaction in both the transparent and/or opaque modes and the last dose distributed over a wide range of non-interactive graphics.

Therefore, a “good” beta value estimate is a key factor for the work of proximity correction under actual conditions. This parameter is extremely sensitive to the composition of the substrate material (in many cases, the electron scattering for statistics/analysis) The exact base definition of the calculation is not possible).

In the satellite window 140, the user has an input section 141 to introduce an estimated line width [μm] value. Selecting "β-Manual" 142 or "β-Auto" 143 in the Accessory Window 140 and starting the process by pressing the button "Start Calculation" 144 will be initiated.

After startup, the user looks for a file of the ASCII-content of the [ ^* .BET] type. The structure of the document requires measurement data from test patterns that also include non-corrected wide lines that have been exposed with various doses.

Fig. 15 shows a table 150 of [ ^* .BET]-documents for preparing a relationship showing the dose [μC/cm ² ] to the exposed line width [μm].

Measurements should be performed on isolated wide lines exposed with various doses. The result of the measurement 160 is shown in Fig. 16. The measured line should be as width > β ("Collect all backscattered electrons", ie for 50KeV masks 10μm) long isolated line exposed graphics. The ( ^* .BET) file (Table 150) is by using an arbitrary content editor, using the "dose" [μC/cm ² ]-separator-"line width" [μm] in the two lines 151,152. Write in the ASCII format as the dose value decreases. This measurement is made using a CD-measuring instrument (eg Leica LMS-IPRO, Leica LWM, CD-SEM, or the like).

This method is based on an analysis of the exposure dose relationship based on line width variation (see Figure 16). The dose is written on the horizontal axis and is increased from the smallest reasonable value by optimal exposure (where the line width is 15 μm and the target meets until the higher value, about 10X to the optimal dose). Fine steps). The measurement line width is shown on the vertical axis. Visual observation of the overall effect of backscattering, along with all additional collisions from the front and back exposure processes, is shown in Figure 16 using the resulting line width and/or format. The specific beta parameter for the structure of a given process can be calculated by introducing a ( ^* .BET)-file into the algorithm under "β-manual" and/or "β-automatic".

Figure 17 shows an ancillary window 170 of the first portion of the main window 120 on the display or for computing the user interface of η. On the PROX-In main window 120, the button "ETA" 171 starts the calculation of "n". In order to achieve this step the previously determined beta values, and two additional experimental values as measurements from isolated transparent long, wide, and narrow lines are required. The calculation of the η-value requires the best dose factor (where the target agrees for both lines): (i) dose to large pattern = _large dose (eg 10 μm or 15 μm exposed line) and (ii) Dosage for small graphics = _small dose (eg 1 μm line), as a non-dimensional value: dose factor = _small dose / _large dose. This value must be introduced into the position titled "Dose Factor" 172, and the exposed graphic size is introduced in the position titled "Graph Size""Large" 173 and "Small" 174 "Width x Length". After selecting the button "ETA" 171 and pressing the button "Start Calculation", the η-value is calculated and displayed in red (see Figure 17).

The second section 180 of the main window 120 is "simulated" and acts as a "fine tuning" of the numerical input parameters on the selected graphic. Parameter tuning is a "back simulation" based on measured dimensions of the graph associated with the applied dose and/or nearby surroundings. There are four types of graphics in terms of disposal, by means of which it is possible to perform a back simulation (Fig. 10). One type of graphic is a single, single transparent line. The relationship between line width variation and exposure exposure is based on the results obtained and the corresponding ( ^* .BET)-file (see Figure 15). Another possibility is the relative optimum dose = the relationship of the dose factor to the line width of the isolated transparent lines, the width ranging from the minimum distinguishable line up to 2-3 μm (according to the process). The ASCII data from the measurement results can be obtained from the linewidth pair ( ^* .TGT)-file dose factor form. The corresponding data can be extracted from the measurement results of the uncorrected exposed test pattern. "PYR" - a cone-like pattern - the change in line width as a function of the gap width of the stylized gap between the gauges along the measurement line and the large and symmetric exposure of the measured lines (see Figure 6). In the ASCII format, measurement data from measurement results in the following form is required: the relationship of the gap width to the line width is taken as ( ^* .PYR)-file. In the ASCII format, the "DRT"-duty factor test from the measurement results in the following form - line width variation as a function of line/space spacing - measurement data is needed: line width versus pitch relationship as ( ^* . DRT) - file. - Data can be extracted from test results on non-corrected test patterns (see Figure 7). The main work in the simulation section, as shown in the second section 180 of the main window 120, is to find (interact) a reasonable set of input parameters for the lithography model used, which simulates the most suitable for the measurement. Good. That means: the simulation should reconstruct the realities of the geometric changes in the measured graphics.

The main work of this simulation part is to find reasonable input parameter values for the lithography model, where the simulation shows the best possible adaptation to the measurement. That means that the simulation should reconstruct the realities of the geometric changes in the measured graphics.

Before pressing the start button 181 for simulation, the user must select one of four graphic types ("LW vs. Q", "To Target...", "PYR", "DRT") from these types. In the test, the corresponding ASCII-file can be obtained. It is also necessary to fill all active Edit-Window 182 (Edit Window 182) with the relevant values and select the desired model method for 2, 3, or 4 Gaussian speed images.

Possible numerical ambiguities (for example, not only single-valued results and/or parameter values without a reasonable physical interpretation) can cause some complex problems. Therefore, we recommend using the "2G" 184 (two Gaussian beam) method to start the simulation and introduce the β- and η-values obtained from the first coarse approximation as the starting value. As the starting value, a value in the range of 0.05 to 0.1 μm can be set.

After starting the analogy, the request corresponding ASCII-file appears [according to the selected graphic type, ( ^* .BET), ( ^* .TGT), ( ^* .PYR), or one of ( ^* .DRT)]. If the file is successfully read and interpreted by the program, the new graphical window 190 (see Figure 19) appears immediately at the top of the second section 180 of the main window 120, which shows the entry from the appropriate graphical form. The value of the measurement 191 and the results of the simulation 192.

The illustration along with the information content appears on the third portion 123 located on the right bottom side of the window 120 (see Figure 12). The third portion 123 contains a corresponding numerical comparison between the experimental results and the calculated evaluation results of the fit quality. In the third part 123, the material can be processed directly, similar to in a normal editor, by marking the content, copying it onto the clip board, and also importing the copied ASCII data directly into other software. Medium (eg Excel, ...) for further processing.

Under "stat" 183, after each analog step on the second section 180, it is indicated that the quality value of the simulation just performed appears. In general, each step in the adaptation process using the back-simulation method to determine the parameters will tend to get the smallest possible value for "stat" 183 (see, for example, the "stat" between Figure 19 and Figure 20 The difference in "the figure 20 is clearly shown to have the characteristics of a better fit 200".

"ind" 193, 203 with an arrow " The form shows the quality tendency to be adapted in the middle of the adaptation process. Pressing the "Set" button 194, 204 sets the current "stat" value to the minimum value for quality assessment, and from now on the indicator "ind" An adaptation quality tendency consistent with this value will be shown. "ind" means: " "-poor;" "-better;" "- No significant changes.

In the case of a selected graphic type, in addition to "DRT", try each of the program functions: "Auto-α", "Auto-β", and "Auto-η" (see Figure 18). It must be possible (check the appropriate window, note: only one can be checked at the same time!). This result is an optimal parameter value recommendation for α or β or η, which appears in red in the second segment 180. If the suggested value is considered reliable, then it should be introduced under the appropriate Edit-Window as the new value for the next simulation step. As the only first step in the auto-adaptation process, it is recommended to start with an "auto-n" function - find an approximately relevant and acceptable value of η. After introducing this value below the edit window, the parameter fit requires many more substitutions of all input parameters.

The window indicated by "3G" and "4G" (see the second section 180 of the main window 120) is for selecting more than two Gaussian parameter set values. Certain areas where measurement results often occur cannot be satisfactorily adapted to simulations using two Gaussian parameter set values using the standard (see Figure 19; range marked with circle 195). Where the measured input of the online wide variation is correct, this may in fact lead to a partial failure of the optimal dose allocation for certain combinations of graphics during the calibration process. In general, it is possible to use multiple two Gaussian functions in order to improve the quality of the adaptation (see Figure 20).

If "3G" or "4G" is used, it is recommended to move two values β and η obtained from the "2G" process as the last two parameters in "3G" or "4G", ie, β γ and η v. This new "blank" parameter β and η should now be set with some "small" starting values and tuned step by step to get the best fit. For fine tuning of α, and β, and η, an "automatic" function is useful.

Figure 21 shows the selection window 210. The user selects "LW vs. Q" 211 and also checks the "Opt.Dose vs. LW" window 212 below. This is a comprehensive control process with a full view of the proposed set of input parameter sets, which does not require any input files. Since the result appears in the comprehensive table 220 of the calculation results from the model (see Fig. 20), it is formed. Table 220 consists of 7 rows and 46 columns. Each column contains the calculated factor for the given line width (from 100 μm to 50 nm exposure) after correction. The line in Fig. 22 has the following meanings: 1-line width [μm] 221, line number: 2 - calculated optimal dose [μC/cm ² ] 222, line number: 3-for a single transparent line Dose factor 223, line number: 4- dose factor 224 for a single transparent contact, line number: 5 - dose factor 225 in the middle line of the L/S (1:1) large array, line number: 6 - in the The exposure density factor 226 from the middle of the (1:1) large array of intermediate lines, and the row number: 7 - the dose factor for the single opaque line 227.

Figure 23 shows the selection window 230. The "to Target..." process allows the user to receive an accurate extraction of the best graphics from a single transparent line used to minimize a given process size. Optimization curve set for the calibration curve.

Figure 24 shows a comparison of the optimal dose of 240 for the measured line width 242 of a single transparent line and a line width 242 of a single transparent line approximated by "2G". The "correction curve" 241 calculated using the "2G" approximation does not provide the best fit to the measured data relative to the line width.

Figure 25 shows a comparison of the optimal dose of 240 for the measured line width of a single transparent line and using a "3G" approximation. The 242 line width of a single transparent line simulated by the "3G" approximation. The "correction curve" 251 calculated using the "3G" approximation provides the best fit of the measured data 252 relative to the line width.

Figure 26 shows a plot 260 of the resulting control function 261 image. The final step in the entire parameter set determination process is the number of control faces 261 used to generate the exposure process optimization. The control function 261 is all determined by the obtained adjacent input parameters α, β, η, .

After checking the "EID to a File ( ^* .pec)" check window, the control of the result in the form of "EID" [Exposure Intensity Distribution] can be obtained for each step in the simulation step. Function 261. This process requires the file name "File Name" (see Figure 24) for the resulting "EID"-file "EID"-File: displayed in the upper part of the display as radial distance [μm] versus exposure density [arbitrary unit A curve image of the control function obtained by the relationship (see Figure 23).

Figure 27 shows the form of a pattern 270 that needs to be written into the resist on the substrate. The graphic 270 is provided by CAD-data and has four different components in this embodiment. The dimensions of the graphic 270 are clearly based on CAD-data or design data. The first component is a straight line 271 having a defined width. The second member 272 is a rectangular shaped region. A line 273 extends from the upper corner and forms the lower corner of the area. The line 271 of the first component is parallel to the line 273 extending from the second component. The third member 274 is a rectangular shaped region. A straight line 273 extends to the left to form a lower corner of the joint. The fourth component 275 includes two regions. These two regions are connected by a straight line 271 at their lower corners. An additional straight line extends from the upper corner of an area to the left. As already mentioned above, all lines 271 are parallel.

The graphics as defined by the CAD-design are divided into sub-shapes 276. Each sub-shape 276 is assigned an electron beam density for the lithography process definition. As a result of program control and the resulting proximity correction, the subshape 276 is further divided and results in an optimal subshape 277. For each individual optimal sub-shape 277, individual doses of the electron beam are dispensed. The dispensing of the dose is carried out according to the optimally adapted parameter set of the correction function.

Figure 28 shows a schematic image written with and without the applied control function. A defined electron beam dose is dispensed to the first and second zones 280 and 281. Schematic pattern caused by electron beam irradiation as shown in the 282 region of the connection between the individual and 284 _{2 2841} 283. According to the CAD-data, it is desirable that the joint areas 284 ₁ and 284 ₂ are separate. Caused by electron beam irradiation in unwanted connections between ₂₈₄₁ and ₂₈₄₂ two lands. Real image arrangement pattern 285 is shown connected between the two lands 300 ₁ and 284 _2. According to the present invention, at least the first and second regions 280 and 281 are divided into two sub-regions 280 ₁ , 280 ₂ , ..., 280 _n and 281 ₁ , 281 ₂ ... and 281 _n , wherein different doses are applied Assigned to a sub-area. According to the present embodiment, the regions 280 and 281 are divided into three sub-regions 280 ₁ , 280 ₂ , and 280 ₃ . Individual doses are assigned to each of the sub-regions 280 ₁ , 280 ₂ , and 280 ₃ , wherein the first sub-region 280 _{1 is} subjected to the dose Do, the second sub-region 280 _{2 is} subjected to the dose D ₁ , and the third sub-region 280 ₃ is subjected to Dose D ₂ . As a result of the various dose assignments for the various sub-regions of the present invention, a structure of the size as required by the CAD data is obtained. A schematic image of various resulting structures 286 shows a clear spacing between the two structures. This spacing is defined by a line 287 having a constant width, and also shows a real image 288 of the graphical structure.

1. . . Electronic lithography system

2. . . Electron beam source

3. . . Electron beam

4. . . Base

5. . . Graphics

6. . . Stage

7. . . electric motor

8. . . electric motor

9. . . Beam alignment coil

10. . . Beam masking unit

11. . . Magnetic deflection unit

12. . . Magnetic coil

13. . . Feedback device

14. . . Electronic detector

15. . . computer

16. . . interface

17. . . Display screen

20. . . Graphics

twenty one. . . region

twenty two. . . Gaussian beam

twenty three. . . Cross section

30. . . Graphics

31. . . Total area

32. . . Formed beam

32 ₁ . . . shape

32 ₂ . . . shape

32 ₃ . . . shape

33. . . Cross section

34. . . arrow

40. . . Polymethyl methacrylate layer

41. . . GaAs substrate

42. . . Simulated orbit

43. . . Electron beam

50. . . Graphics

51. . . First part

52. . . Second part

53. . . Third part

54. . . Fourth part

55. . . Get the graphics

56. . . First position

57. . . Second position

60. . . First test pattern

61. . . line

62. . . Flange

63. . . Line width

64. . . Gap width

65. . . point

70. . . Second possible test pattern

71. . . Line width

72. . . line

73. . . spacing

74. . . Array

75. . . line

76. . . Single transparent line

80. . . Input window

81. . . mark

90. . . table

91. . . first row

92. . . second line

93. . . The third row

94. . . Curve form

95. . . Parameter group

96. . . Measurement data

97. . . Calculation data

100. . . Input window

101. . . mark

110. . . table

111. . . first row

112. . . second line

113. . . The third row

114. . . Graphical form

115. . . Parameter group

116. . . Measurement data

117. . . Calculation data

120. . . Main window

121. . . first part

121 ₁ . . . Subsidiary box

121 ₂ . . . Subsidiary box

121 ₃ . . . Subsidiary box

121 ₄ . . . Subsidiary box

122. . . the second part

123. . . the third part

124. . . Content window

130. . . Accessory window

131. . . Select button "ALPHA"

132. . . Input section

133. . . Input section

134. . . Input section

140. . . Accessory window

141. . . Input section

142. . . Button "β-manual"

143. . . Button "β-automatic"

144. . . Button "Start calculation"

150. . . table

151. . . Row

152. . . Row

160. . . result

170. . . Accessory window

171. . . Button "ETA"

172. . . "dose factor"

173. . . "Graphic size" "large"

174. . . "small"

180. . . Second section

181. . . Start button

182. . . Edit window

183. . . "stat"

184. . . "2G"

190. . . Graphical window

191. . . Measurement result

192. . . simulation

193. . . "ind"

194. . . "set" button

195. . . Virtual circle

203. . . "ind"

204. . . "set" button

210. . . Select window

211. . . "LW vs Q"

212. . . Windows

220. . . table

221. . . Line width

222. . . Calculated optimal dose

223. . . Dose factor

224. . . Dose factor

225. . . Dose factor

226. . . Exposure density factor

227. . . Single opaque line

230. . . Select window

240. . . Comparison

241. . . Calibration curve

242. . . Line width

250. . . Comparison

251. . . Calibration curve

252. . . data

260. . . curve

261. . . Control function

270. . . Graphics

271. . . straight line

272. . . the second part

273. . . straight line

274. . . the third part

275. . . fourth part

276. . . Sub shape

277. . . Sub shape

280. . . First district

280 ₁ . . . Subregion

280 ₂ . . . Subregion

280 ₃ . . . Subregion

281. . . Second district

281 ₁ . . . Subregion

281 ₂ . . . Subregion

281 ₃ . . . Subregion

282. . . Schematic image

283. . . Junction

284 ₁ . . . Individual junction

284 ₂ . . . Individual junction

285. . . Real image

286. . . Result structure

287. . . straight line

288. . . Real image

The characteristics and modes of operation of the present invention will now be described more fully in the following detailed description of the invention, in which: FIG. 1 is a block diagram of an electron beam lithography system; and FIG. 2a is a pair of Gaussian beams An example of a written image; a 2b image is a cross-sectional shape of a Gaussian beam having a constant diameter; a 3a is an example of a pattern with a shaped beam writing; and a 3B is a cross-sectional shape of the shaped beam, wherein The shape of the beam can be adjusted according to the pattern to be written; Figure 4a shows the simulated orbit of 100 electrons scattered in polymethyl methacrylate (PMMA) plated on a GaAs substrate; Figure 4b shows An analog orbit of 100 electrons scattered in polymethyl methacrylate (PMMA) plated on a GaAs substrate, wherein the initial energy of the electron is relatively high as shown in the calculation of Fig. 4a; Fig. 5a A schematic diagram showing the pattern of the pattern to be written in the resist on the substrate; Figure 5b shows the result of the pattern, the pattern being written in the resist, and the correction according to the invention is not applied; Figure 6 shows the write to the resist a first test pattern within; FIG. 7 shows a second test pattern written into the resist; FIG. 8 shows the user's input window to begin with the first shown in FIG. Exposure of the test pattern; FIG. 9 shows a measurement result table obtained from the exposed first test pattern; FIG. 10 shows an exposure for the user input window to start the second test pattern shown in FIG. 7; Fig. 11 shows a measurement result table obtained from the exposed second test pattern; Fig. 12 shows a main window of the program PROX-In provided on the display screen associated with the computer; Fig. 13 shows the display a sub-window of the first portion of the main window of the user interface on the user interface for calculating alpha; Figure 14 shows a sub-window of the first portion of the main window on the display screen or the user interface for calculating β and η; Figure 15 shows a table of the exposed line width as a function of the applied dose; Figure 16 shows the change in the line width as a function of the exposed reagent; Figure 17 shows the use in the display screen or for the calculation of η The child window of the first part of the main window; Figure 18 shows the main The second section of the window (see Figure 12), which acts as a "fine tuning" of the input parameters numerically represented; Figure 19 shows a simulation of the 15 μm linewidth variation using the "2G" approximation; The figure shows a simulation of a 15 μm line width variation using a "3G" approximation; a 21st figure showing a selection window; a 22nd figure showing a comprehensive table with calculation results from the model; and a 23rd drawing showing a selection window; The comparison shows the best dose simulation comparison of the measured and simulated linewidths of a single transparent line, using a "2G" approximation; the 25th plot compares the measured and simulated linewidths of a single transparent line. The optimal dose comparison uses a "3G" approximation; Figure 26 shows a graphical representation of the resulting control function; Figure 27 shows a schematic representation of the graph to be written by the lithography process, and the graph The collapse into the sub-element, and Figure 28 shows a schematic of the graph written with and without the applied control function.

1. . . Electronic lithography system

2. . . Electron beam source

3. . . Electron beam

4. . . Base

5. . . Graphics

6. . . Load

7. . . electric motor

8. . . electric motor

9. . . Beam alignment coil

10. . . Beam masking unit

11. . . Magnetic bias unit

12. . . Magnetic coil

13. . . Feedback device

14. . . Electronic detector

15. . . computer

16. . . interface

17. . . Display screen

Claims (17)

A method for controlling proximity effect correction in an electron beam lithography system, wherein an exposure controlled by a control function is to obtain a final obtained pattern consistent with design data after processing, the method comprising the following steps: Exposing any set of graphics without applying the method for controlling proximity correction; * measuring the geometry of the resulting test structure and obtaining a set of measurements; * determining the basis from the data set of the measurement Entering adjacent parameters; * applying the control function to the exposure control of the electron beam lithography system during pattern exposure according to the design data; * changing at least the basic input proximity parameters of the control function; and adapting a model to the measurement The resulting data set, and thus the optimized set of neighboring parameters; * the separately changed neighboring parameters are applied to the calculation, and the calculated result is compared to the data set of the measured result. The method of claim 1, wherein the method comprises the step of applying a determined set of adjacent parameters to the calculation and comparing the results to measurements of isolated transparent and opaque lines exposed with normal doses. The method of claim 1, wherein the method comprises adapting The proximity parameter set is applied to the calculation, and the results are compared with measurements from other arbitrary conical-like graphs, and the results are compared with the measurement data set from the measurement results at the representative points of the test pattern. . The method of claim 1, wherein the method comprises applying the adapted set of neighboring parameters to a calculation, and the measurement data set of the result and the graph of the plurality of lines from the other arbitrary duty factor A comparison and comparison of these results with measurements from measurements at representative points of the test pattern. The method of claim 1, wherein the control function is a sum of at least two Gaussian functions. The method of claim 5, wherein the control function is constituted by Determined, wherein the first term α- denotes a short-range feature of forward scatter, the second term β-backscatter, the parameter η is the ratio of the backscattered component to the deposition energy of the forward scatter component, and γ- is The distance from the point of incidence of the electron. The method of claim 1, wherein the control function is determined by a sum of three Gaussian functions. The method of claim 1, wherein the control function is determined by a sum of four Gaussian functions. The method of claim 1, wherein the graphic according to the design data is subdivided into optimal sub-shapes, and individual doses are assigned to each The best sub-shape, wherein the individual dose is determined from the correction function. The method of claim 9, wherein the individual doses assigned to the optimal sub-shape result in an exposed pattern that is in the same dimensions as the design data. The method of any one of claims 1 to 10, wherein the proximity correction control function is utilized in an electron beam lithography system to reduce the size error to 10 nm for a 100 nm device generation (100 nm device generation) . The method of claim 1, wherein the proximity control function is embedded in a software of PROX-In. The method of claim 12, wherein the PROx-In software operates on a standard computer connected to the electron beam lithography system, connects the display screen to the computer, and displays the user as a user on the display screen. The main window (120) of the interface, wherein the main window appears immediately after starting the program PROX-In, and the main window is divided into three main parts. The method of claim 13 wherein the first portion of the main window is comprised of first, second, third and fourth separate sub-windows, and wherein the separate sub-windows are used to calculate the parameters . The method of claim 13, wherein the second portion of the main window acts as a "fine tuning" of the numerical input parameter set on the selected test pattern, wherein the parameter tuning is dependent on the applied The "back simulation" of the measured size change of the dose and / or the vicinity of the graph. For example, the method described in claim 15 can select four types. The graphics are used to perform the back simulation, the first graphic is a wide single transparent line, and the relationship of the resulting line width change to the exposure dose is based on the resulting result and the corresponding (*.BET)-file that occurs, The second test pattern is an isolated transparent line which should have a finite width, where the relative optimum dose = dose factor versus isolated transparent line width, the width ranging from the smallest distinguishable line up to 2-3 μm line It is exposed, and the ASCII-data is obtained from the relationship of the dose factor from the (*.TGT)-file from the measurement result in the line width manner; the third test pattern is "PYR"-like tapered pattern - wherein the line width The change is obtained as a function of the programmed gap width between the measured line and the large, symmetrically exposed pad along the measured line, and the data of the measurement is in ASCII-format Medium, and the same as the gap width to line width (*.PYR)-file to display; the fourth test pattern is "DRT" - duty factor test line width change as a function of line / interval - spacing - measurement results And the information is in ASCII- In the format, and the line width is the same as the spacing (*.DRT)-file, all data is extracted from the measurement results on the uncorrected test pattern. The method of claim 13, wherein the third portion of the main window is located in a right bottom side of the main window, and the third portion is included between the experimental data and the calculation result with the adaptation quality evaluation. Numerical comparison, in which the data can be processed directly in the third part as in the normal editor, that is, the content is marked, copied to the cut version, and the copied ASCII data is directly introduced into other software ( For example, Excel, ...) is used for further processing.